Group 4 Project Presentation: Gene Expression in bone marrow derived cells in healthy donors and patients with myelodysplastic syndrome

-Tilman Nelissen (s233043) -Thea Rehm Rosholm (s225019) -Edir Sebastian Vidal Castro (s243564) -Nithiyashri Jayashankar (s244356) -Morten Orebo Holmström (s242223)

2024-12-03

Introduction

  • The myelodysplastic syndromes (MDS) are a diverse set of cancers in the bone marrow.
  • The bone marrow consists of several cell types - amongst these hematopoietic stem cells and progenitor cells.
  • Gene expression analysis of specific cell fractions derived from the bone marrow could reveal potential targets of therapy and immune evasive mechanisms.

From: https://www.sigmaaldrich.com/

Materials and methods: Data Retrieval and Cleaning

Flowchart of data aquisition and inital tidying

Methods: Gene Mapping, Cleanup and DESeq2 Analysis

Flowchart of data augmentation

Sample compositions

(A) Overall cellular composition of the MDS patient samples as well as healthy donors

(A)

(B)

(B) The following plots depicts the total number as well as the fraction of MDS and HD samples that contain the srsf2 and sf3b1 mutations within the various cellular compartments.

Distribution of counts

Histogram non-normalized

Histogram normalized

The histograms above illustrate gene counts from both MDS patients and healthy donors, displayed before and after normalization.

The impact of normalization appears subtle. However, it is noticeable that the overall gene counts in MDS patients are comparatively higher than those in healthy donors.

Clustering using PCA on all cells and on HSC

Clustering analysis using K-means (k = 4) compared to umap:

Differential gene expression: All samples and HSC

Differential gene expression across the different cell populations - MDS vs. HD

Conclusion and perspectives

  • Several genes are differentially expressed between healthy donors and patients with MDS in the four analyzed cell populations.

  • Single cell analysis (RNASeq or proteomics) would provide further detail.

  • Differences in expression between the SF3B1/SRSF2 mutants and wild types?

  • More sample/clinical data on MDS samples would allow to analyze these more in deeply and gain more knowledge of differential gene expression between different classes of MDS.

22160 - R for Bio Data Science